Publication number | US7603306 B2 |
Publication type | Grant |
Application number | US 11/225,219 |
Publication date | Oct 13, 2009 |
Filing date | Sep 13, 2005 |
Priority date | Jul 30, 1999 |
Fee status | Paid |
Also published as | CA2314628A1, US6947904, US7962388, US7962392, US8566196, US20060010061, US20080243665, US20100076905, US20110276517 |
Publication number | 11225219, 225219, US 7603306 B2, US 7603306B2, US-B2-7603306, US7603306 B2, US7603306B2 |
Inventors | Rex Macey |
Original Assignee | Rex Macey |
Export Citation | BiBTeX, EndNote, RefMan |
Patent Citations (22), Non-Patent Citations (1), Referenced by (9), Classifications (15), Legal Events (3) | |
External Links: USPTO, USPTO Assignment, Espacenet | |
This application is a continuation of U.S. application Ser. No. 09/625,702 filed Jul. 25, 2000 (now U.S. Pat No. 6,947,904, which issued Sep. 20, 2005), which claims priority benefit of Provisional Application Nos. 60/146,532; 60/168,331, which were filed on Jul. 30, 1999 and Dec. 1, 1999 respectively. The aforementioned applications are hereby incorporated herein by reference in their entireties.
1. Field of the Invention
The present invention relates to financial management systems and, more particularly, to a method and an apparatus for effecting a retirement planning system incorporating means for planning, analyzing, and preparing reports on an investment portfolio comprised of various financial assets owned by one or more investors.
2. Description of the Related Art
Retirement planning is a frustrating exercise because the future is uncertain. Nonetheless, prior art retirement calculators often employ deterministic models that output seemingly precise predicted value of a portfolio. Typically, a user is required to input a number of parameters such, for example, as a fixed rate of inflation, a fixed rate of return, and a specified retirement age. Using these user-input values, the prior art calculators compute a future value of the portfolio at the specified retirement age or at some other specified point in time. This computed portfolio value, however, is of limited use to an investor, or may even be misleading to the investor, because the chance of achieving the computed value is not at all certain, but lies somewhere between 0% and 100%. As any competent investor knows, future rate of return or rate of inflation varies from month-to-month and from year-to-year, depending on a myriad of highly variable economic parameters. Moreover, one's actual life span may be shorter or longer than that published in a life expectancy table, thereby creating further uncertainties in the computed portfolio value.
U.S. Pat. No. 5,918,217 to Maggioncalda et al. discloses a financial advisory system for assisting an investor to select a mix of financial products that would achieve a specified retirement goal. The system has a user interface that enables the investor to interactively explore how changes in one or more input decisions such as risk tolerance, savings level, and retirement age affect one or more output values such as a probability of achieving a financial goal. The system further includes a simulation module that generates a probability distribution for a projected asset valuation. However, Maggioncalda et al. does not disclose, for example, the use of Monte Carlo simulation techniques to vary the retirement age of the investor to more realistically demonstrate the chances of achieving the investor's retirement goals.
Accordingly, there is a need for a retirement planning system that uses Monte Carlo techniques to vary the life span of an investor and/or to vary the rate of return of a portfolio so as to more completely describe an investor's chances of achieving a retirement goal.
An object of the present invention is to provide an easy-to-use and a cost-effective method and apparatus for effecting an improved financial management system that enables a user (e.g. an investor) to realistically predict or project the future value of an investment portfolio while taking into consideration variations in the rate of return of the investment portfolio and/or mortality risks of the investor(s) over time.
According to one aspect of the invention, the inventive system computes possible future values of a portfolio while varying an investor's life span and the portfolio's rate of return using Monte Carlo simulation techniques. The portfolio may consist of assets owned by a single investor or jointly by a plurality of investors.
According to another aspect of the invention, the inventive system randomly generates a rate of return for a subdivided time period T_{s}, such for example, as a year or a month, from a user-supplied distribution of rates of return such as historical returns or a lognormal distribution having a specified mean and standard deviation.
According to still another aspect of the invention, the inventive system produces possible future returns or values of an investment portfolio by varying hypothetical life spans of the investor over a plurality of trials. The system varies the length of each trial by randomly (or pseudo-randomly) generating a number between (and including) 0 and 1 in each trial for each investor and comparing the generated number with a corresponding probability of dying (i.e. mortality rate) of each investor using a mortality table. An investor's hypothetical life in a trial is terminated if his/her mortality rate is greater than or equal to the randomly generated number; otherwise, the investor lives to an age of death indicated by the mortality table in accordance with the generated number. A trial ends when the hypothetical lives of all joint investors are terminated in the trial. Optionally, a random (or pseudo-random) number is generated for each investor at the beginning of each subdivided period T_{s}, and which is then compared to a chance of dying from the mortality table so as to determine whether the investor lives the upcoming subdivided period T_{s}.
According to yet another aspect of the invention, the inventive system employs user-specified mortality rates or mortality factors (e.g. age, sex, and race) to modify published mortality rates so as to more accurately model an investor's anticipated life span.
According to still another aspect of the invention, the inventive system computes net cash flow of a portfolio, while taking into account scheduled contributions and withdrawals by the investor(s) during a specified period.
The inventive system advantageously produces useful results such as the following:
1. the value of a portfolio during each subdivided period in each trial;
2. the value of a portfolio at the end of each trial;
3. whether a portfolio, in each trial, has met a stated investment objective;
4. the probability of meeting (or not meeting) an investment objective for all trials;
5. the distribution of portfolio values over a specified period for each trial and the remainder value of the portfolio after all specified trials are completed;
6. the age-at-death of each investor at the end of each trial;
7. the distribution of each investor's age-at-death for all trials;
8. the distribution of ages of the investors in the group for all trials; and
9. the probability of each investor living to a specified age.
In one embodiment, there is provided a retirement planning method for computing possible future values of a portfolio of an investor. The method includes the steps of (a) receiving a plurality of user inputs comprising an initial value of the portfolio and a current age of the investor; (b) providing data indicating cumulative probabilities of living to an age of death and/or cumulative probabilities dying at an age of death for persons of a given age group; (c) randomly drawing a number between and including 0 and 1 for the investor; (d) defining the randomly drawn number as either a cumulative probability of living to an age of death or a cumulative probability of dying at an age of death; (e) determining an age of death of the investor in accordance with the data based on the current age of the investor and the randomly drawn number; (f) computing a future value of the portfolio using the age of death of the investor determined in step (e), a predetermined rate of return, and the initial value of the portfolio; and (g) outputting the computed future value of the portfolio.
In another embodiment, a retirement planning method for computing a possible future value of a portfolio of an investor includes the steps of:
(a) receiving user inputs comprising an initial value of the portfolio and a current age of the investor;
(b) randomly drawing a number between 0 and 1 (which may include one or more end points);
(c) determining a mortality rate of the investor in accordance with a mortality table based on the current age of the investor;
(d) comparing the randomly drawn number with the determined mortality rate using a preselected logical relation:
(e) computing a future value of the portfolio using the age of death defined in step (d)(1), a predetermined rate of return, and the initial value of the portfolio; and
(f) outputting the computed future value of the portfolio.
In yet another embodiment, a retirement planning method for computing a possible future value of a portfolio of an investor, comprises the steps of:
(a) receiving user inputs comprising an initial value of the portfolio and a current age of the investor;
(b) randomly drawing a number between 0 and 1 (which may include one or more end points) for the investor;
(c) determining a mortality rate of the investor in accordance with a mortality table based on the current age of the investor;
(d) comparing the randomly drawn number with said determined mortality rate using a preselected logical relation such that:
In still another embodiment, a retirement planning method for computing a possible future value of a portfolio of a plurality of joint investors, includes the steps of:
(a) receiving user inputs comprising an initial value of the portfolio, a current age of a first joint investor, and a current age of a second joint investor;
(b) randomly drawing a number between 0 and 1 (which may include one or more end points) for each of the first and second joint investors;
(c) determining a mortality rate of each of the first and second joint investors in accordance with a mortality table based on the current age of the first and second joint investors;
(d) comparing the randomly drawn numbers of the first and second joint investors with corresponding mortality rates determined for the first and second joint investors in step (c) using a preselected logical relation such that:
Definitions or descriptions of some key terms are as follows:
Mortality table: a table containing the probability that a person who reaches a specified age dies within the specified period. For example, a table may show that a person who reaches the age of 50 has a 0.30% chance of dying while age 50.
Vitality table: a table containing the probability that a person who reaches a specified age lives through the specified period. These probabilities are equal to 100% less the corresponding probability in a mortality table.
Mortality table with Cumulative Probabilities: a table containing the probability that a person will not live to a specified age (i.e. die before the specified age). For example, such a table may show that a person has a 40% probability of not reaching age 75 (dying before age 75). This type of table may be derived from the mortality table.
Vitality table with Cumulative Probabilities: a table containing the probability that a person will live to a specified age. For example, such a table may show that a person has a 60% probability of reaching age 75. This person will have a higher probability of reaching younger ages, and a lower probability of reaching higher ages. This type of table may be derived from the mortality table.
Joint mortality table: a table containing the probability that Person 1 will die at Age X and Person 2 will die at Age Y. Such a table may be constructed from mortality table for each person. A mortality table for a Person 1 would have probabilities of dying at ages X+1 through age 110 (or some other specified age). Such a table would contain 110−X entries. A joint mortality table would have probabilities for every possible pair of ages at death. For a table up to age 110, this would be [(110−X)*(110−Y)] probabilities. From our perspective, the two tables both contain probabilities that an event will happen. The mortality table gives the probability that a person dies at a given age. The joint mortality table gives the probability that two persons die at two specified ages. The joint mortality table can be used to construct a joint mortality table with cumulative probabilities.
Other objects and features of the present invention will become apparent from the following detailed description considered in conjunction with the accompanying drawings. It is to be understood, however, that the drawings are designed solely for purposes of illustration and not as a definition of the limits of the invention, for which reference should be made to the appended claims. It should be further understood that the drawings are not necessarily drawn to scale and that, unless otherwise indicated, they are merely intended to conceptually illustrate the structures and procedures described herein.
In the drawings, wherein like reference characters denote similar elements and method steps throughout the Figures:
The present invention is directed to a data processing system for use by a financial advisor or investor(s) for investment planning such, for example, as retirement planning. The inventive data processing system is configured to apply Monte Carlo simulation techniques to compute or predict possible future values of one or more investment portfolios while varying an investor's life span and/or rates of return during each hypothetical life span of an investor. A portfolio may be owned entirely by a single investor or jointly by a group of investors. The possible future values of the investment portfolio are calculated for a number of trials (e.g. 1000), with each trial representing a possible (or hypothetical) life of the single investor or a “joint” life of the group of investors (e.g. a married couple). The system may also take into account cash flow including, without limitation, contributions by an investor during his/her working years and withdrawals by the investor during his/her retirement years.
In a preferred embodiment, the system randomly or pseudo-randomly generates a rate of return from data representing historical returns or a user-specified distribution such, for example, as a lognormal distribution having a specified mean and standard deviation. (It is noted that the terms “random” and “pseudo-random” are used interchangeably herein.) The system also randomly or pseudo-randomly generates a number between 0 and 1 (e.g. a number greater than zero and less than one and which may include one or more end points) for each investor and compares the randomly generated number with another number representing the investor's chance of dying (from, for example, a mortality table) at the beginning of each trial, for use in determining whether the investor lives in that trial and, if so, for how long. The mortality table may be based on published mortality rates that preferably take into account such mortality factors as, for example, gender, race, and sex. Alternatively, the mortality table may include or comprise mortality rates specified by the user. Instead of using a mortality table, the system includes methods for randomly generating ages at death, as explained more fully below, or the system may assume a fixed life span, as specified by the user, for each investor for all trials. In the latter case, only the rates of return are varied during each trial.
Referring now to the drawings,
The CPU 14 typically comprises, without limitation, arithmetic logic units AU 22, AU 24, and AU 26. AU 22 is configured to store and/or retrieve data from the database(s) stored in data storage device 12. These database(s) may include information such, for example, as historical return data for various asset classes or investment instruments (e.g. long-term corporate and government bonds, or large cap stocks), and a distribution of returns such as a lognormal distribution with a user-specified mean (e.g. 8%) and a user-specified standard deviation (e.g. 9%). The database(s) may further include mortality rates, published and/or user-specified, and mortality factors based on age, sex, race, health, etc. for modifying the mortality rates. Instead of or in addition to these mortality rates, database(s) may also include cumulative probabilities of living to an age of death and/or cumulative probabilities of dying at an age of death. AU 24 is configured to manipulate the retrieved data by, for example, calculating incremental increases and decreases of an investment portfolio based on a rate of return. AU 26 is configured to generate a random or a pseudo-random number for comparison with, for example, another number representing the mortality rate of an investor, the purpose of which comparison will become apparent as this disclosure proceeds.
As seen in
If the response to the block 106 query is positive, the method requests at block 108 (
As seen at block 126 in
Where the answer to the query of block 106 (
At block 138, the current value of L_{MI }is compared to the current value of L_{RI }according to a preselected logical relation such, for example, as “greater than or equal to” (i.e. “≧”). For example, if it is determined that L_{MI }is not greater than or equal to L_{RI }for all investors, then the method proceeds to block 140 (
The computed results may be displayed on a monitor or printed in a tabulated and/or graphical form. Based on these results, the user may revise user-specified parameters such as the average and standard deviation of the rate-of-return distribution and/or the age-at-death of each investor and then the have system execute another simulation run. The user can thus run multiple simulations until acceptable results have been obtained.
As depicted in
If a fixed life span is specified at block 206, then the user must specify (at block 208) the age-at-death of each investor. At block 210, the system discretizes or divides time into a plurality of subdivided periods (as for example by defining units of time in years, months etc.) so that growth of the investment portfolios may be expressed in terms of the subdivided periods. At block 212, trial counter J is initialized for counting the number of trials prior to a simulation run.
At block 214, the method queries whether the current age of each investor exceeds his/her specified age-at-death. If the answer is negative, the method proceeds to block 216 (
Where the answer to the query of block 206 (
In accordance with a particularly advantageous embodiment of the invention, the method determines the age-at-death of an investor by generating a random (or pseudo-random) number L_{RI }for each subsequent subdivided period and compares that number with a corresponding mortality rate L_{MI }for that subdivided period. The method continues to this comparison process while updating the current age of the investor at each subdivided period until the hypothetical life of the investor ends, i.e. when the L_{RI }generated for a subdivided period is less than or equal to the L_{MI }for that subdivided period. At such time, the method records the age of the investor during the last subdivided period as the age-at-death for that investor.
At block 250 (
The query at block 268 then determines whether the specified N trials have been completed by asking if the current value of J is greater than N? If so, the simulation run is ended at block 270; otherwise, the method at block 272 resets parameters such as the current age of all investors, the trial counter J, and the subdivided period T_{s }to their initial values and proceeds to block 242 (
At the end of a simulation run, the user may display the results of that simulation run and determine or decide whether another simulation run is desired. If so, the user may input different initial values for parameters such as inflation rate, mean and standard deviation for a rate-of-return distribution, savings and withdrawals, etc.
In another embodiment, there is provided a retirement planning method for computing possible future values of a portfolio of an investor, which includes the steps of (a) receiving a plurality of user inputs comprising an initial value of the portfolio and a current age of the investor; (b) providing data indicating cumulative probabilities of living to an age of death and/or cumulative probabilities dying at an age of death for persons of a given age group; (c) randomly drawing a number between and including 0 and 1 for the investor; (d) defining the randomly drawn number as either a cumulative probability of living to an age of death or a cumulative probability of dying at an age of death; (e) determining an age of death of the investor in accordance with the data based on the current age of the investor and the randomly drawn number; (f) computing a future value of the portfolio using the age of death of the investor determined in step (e), a predetermined rate of return, and the initial value of the portfolio; and (g) outputting the computed future value of the portfolio.
In still another embodiment, a retirement planning method for computing a possible future value of a portfolio of a plurality of joint investors, includes the steps of:
(a) receiving user inputs comprising an initial value of the portfolio, a current age of a first joint investor, and a current age of a second joint investor;
(b) providing data indicating one of cumulative probabilities of living to an age of death and cumulative probabilities dying at an age of death for persons of a given age group;
(c) randomly drawing a number between 0 and 1 (which may include one or more end points) for the first joint investor;
(d) defining the randomly drawn number of step (c) as one of said one of cumulative probabilities of living to an age of death and cumulative probabilities of dying at an age of death for the first joint investor;
(e) determining an age of death of the first joint investor in accordance with said data, based on the current age of the first joint investor and the randomly drawn number of step (c);
(f) randomly drawing a number between 0 and 1 (which may include one or more end points) for the second joint investor;
(g) defining the randomly drawn number of step (f) as one of said one of cumulative probabilities of living to an age of death and cumulative probabilities of dying at an age of death for the second joint investor;
(h) determining an age of death of the second joint investor in accordance with said data based on the current age of the second joint investor and the randomly drawn number of step (f);
(i) determining the greater age of death of the first and second joint investors by comparing the age of death of the first joint investor determined in step (e) with the age of death of the second joint investor determined in step (h); and
(j) computing a future value of the portfolio using said greater age of death of the joint investors, a predetermined rate of return, and the initial value of the portfolio; and
(k) outputting the computed future value of the portfolio.
In still another embodiment, a retirement planning method for computing a possible future value of a portfolio of an investor, includes the steps of (a) receiving user inputs comprising an initial value of the portfolio and a current age of the investor; (b) randomly drawing a number between 0 and 1 (which may include one or more of the end points); (c) determining a mortality rate of the investor in accordance with a mortality table based on the current age of the investor; (d) comparing the randomly drawn number with the determined mortality rate using a preselected logical relation such, for example, as a logical operator representing “less than or equal to” (i.e. “≦”) such that:
(1) if the randomly drawn number satisfies the preselected logical relation with the determined mortality rate of the investor, define the current age as the age of death of the investor;
(2) if the randomly drawn number does not satisfy the preselected logical relation with the determined mortality rate, advance the current age of the investor to a next age group indicated in the mortality table and repeat steps (b) through (d);
(e) computing a future value of the portfolio using the age of death defined in step (d)(1), a predetermined rate of return, and the initial value of the portfolio; and
(f) outputting the computed future value of the portfolio.
In still another embodiment, a retirement planning method for computing a possible future value of a portfolio of an investor, includes the steps of:
(a) receiving user inputs comprising an initial value of the portfolio and a current age of the investor;
(b) randomly drawing a number between 0 and 1 (which may include one or more end points);
(c) determining a mortality rate of the investor in accordance with a mortality table based on the current age of the investor;
(d) comparing the randomly drawn number with said determined mortality rate using a preselected logical relation such that:
(e) computing a future value of the portfolio using the age of death defined in step (d)(1), a predetermined rate of return, and the initial value of the portfolio; and
(f) outputting the computed future value of the portfolio.
In yet another embodiment, a retirement planning method for computing a possible future value of a portfolio of a plurality of joint investors, includes the steps of:
(a) receiving user inputs comprising an initial value of the portfolio, a current age of a first joint investor, and a current age of a second joint investor;
(b) randomly drawing a number between 0 and 1 (which may include one or more end points) for the first joint investor;
(c) determining a mortality rate of the first joint investor in accordance with a mortality table based on the current age of the first joint investor;
(d) comparing the randomly drawn number with said determined mortality rate for the first joint investor using a preselected logical relation such that:
(e) randomly drawing a number between 0 and 1 (which may include one or more end points) for the second joint investor;
(f) determining a mortality rate of the second joint investor in accordance with the mortality table based on the current age of the second joint investor;
(g) comparing the randomly drawn number with said determined mortality rate for the second joint investor using the preselected logical relation such that:
(h) determining the greater age of death of the joint investors by comparing the age of death of the first joint investor defined in step (d)(2) with the age of death of the second joint investor defined in step (g)(2);
(i) computing a future value of the portfolio using the greater age of death determined in step (g), a predetermined rate of return, and the initial value of the portfolio; and
(j) outputting the computed future value of the portfolio.
The random numbers may be drawn so as to provide greater assurance that the ages of death derived for the investor over the number of trials are distributed in a manner similar to that implied by the mortality (or vitality) table. In one embodiment, the interval between 0 and 1 is divided into a plurality of partitions (e.g., m number of partitions), and a preselected quantity (S_{i}) of random numbers is drawn from each of the plurality of partitions. The preselected quantity (S_{i}) of random numbers is preferably proportional to the size of the corresponding partition. Thus, the interval of 0 to 1 may, for example, be divided into three partitions: 0 to 0.1, 0.1 to 0.5, and 0.5 to 1.0 (representing 10%, 40%, and 50% of the interval, respectively). For a sample size of 1000 random numbers, 100 random numbers may be drawn from the partition 0 to 0.1 (the first partition), 400 random numbers from 0.1 to 0.5 (the second partition), and 500 from 0.5 to 1.0 (the third partition). The random numbers from each partition may or may not include the end points of each partition. Each partition may or may not be of the same size.
An example is illustrated as follows: For each partition from i=1 to m, select S_{i }random numbers between end points LBOUND and UBOUND where LBOUND=(Sum of S_{j})/N for j=1 to i−1 (note that LBOUND=0 when i=1), UBOUND=(Sum of S_{j})/N for j=1 to i, and N=the total number of random numbers. Generate an Age At Death of an investor using each random number in place of the random drawn number called for in any one of the methods described above.
More specifically:
Step 1. Select m as the number of partitions and S(i) as the quantity of random numbers to select from the i^{th }partition.
Step 2. Let N=Sum of S(i) for i=1 to m
Step 3. Let PARTITION_CTR=1
Step 4. Let LBOUND=0
Step 5. Let UBOUND=S(i)
Step 6. Let CUM_SI═S(PARTITION_CTR)
Step 7. Let j=1
Step 8. Let RND=a random number between 0 and 1
Step 9. Let RND=(RND*(UBOUND−LBOUND)+LBOUND)/N
Step 10. Generate an Age At Death of an investor using one of the above-described methods using RND in place of the random number.
Step 11. Let j=j+1
Step 12. If j is less than or equal to S(PARTITION_CTR) then goto Step 8, otherwise continue
Step 13. Add 1 to PARTITION_CTR
Step 14. if PARTITION_CTR>m then Stop otherwise continue
Step 15. Let CUM_SI═CUM_SI+S(PARTITION_CTR)
Step 16. Let LBOUND=UBOUND
Step 17. Let UBOUND=CUM_SI
Step 18. Goto Step 7
In another embodiment, the interval of 0 to 1 is divided into a sequence of numbers uniformly distributed between 0 and 1, for example: 0.1, 0.2, 0.3 . . . 0.9. The end points (i.e., 0 and 1) may or may not be included in the sequence. Each of these uniformly distributed numbers may then be used in place of the randomly drawn numbers as described in the above embodiments. For example, each of these uniformly distributed numbers may be used in conjunction with a mortality (or vitality) table to determine the age of death of an investor for a subdivided period or a trial. Note that this method produces N ages at death distributed as the mortality (or vitality) table would imply.
A specific example is illustrated as follows:
Step 1. Start. Select a sample size of N (e.g. 1000).
Step 2. Let a variable RND=0
Step 3. Add 1/(N+1) to RND
Step 4. Generate an Age At Death using any of the above-described methods using RND in place of the randomly drawn number.
Step 5. If RND>=N/(N+1) then stop. Otherwise return to Step 3.
As stated above, as an alternative to using the mortality tables to select a life expectancy of an investor, ages at death can be derived from a mortality table, a vitality table, a mortality table with cumulative probabilities, or a vitality table with cumulative probabilities. Moreover, such ages can be randomly generated by the following inventive techniques, which can, alternatively, be used to generate an age of death for use in the embodiments disclosed above.
Method 1. Using Mortality Table
Step 1. Start. Let the Age at Death equal the current age of the person.
Step 2. Increment the Age At Death by one time period (e.g. by one year).
Step 3. Look up probability of dying at the Age at Death in this period in the mortality table.
Step 4. Generate a random number uniformly distributed between 0 and 1. If the random number is less than the probability of death, then stop. Otherwise return to Step 2.
Method 2. Using Vitality Table
Step 1. Start. Let the Age At Death equal the current age of the person.
Step 2. Increment the Age At Death by one time period (e.g. by one year).
Step 3. Look up probability of living through this period in the vitality table.
Step 4. Generate a random number uniformly distributed between 0 and 1. If the random number is greater than or equal to the probability of living, then stop. Otherwise return to Step 2.
Method 3. Using Mortality Table with Cumulative Probabilities
Step 1. Start.
Step 2. Generate a random number uniformly distributed between 0 and 1.
Step 3. Using the Mortality table with the Cumulative Probabilities table, find the age which most closely corresponds to the probability generated in Step 2. This age is the Age at Death.
Method 3a. Using Mortality Table (More General Case of Method 3)
Step 1. Start
Step 2. Divide the segment 0-1 into intervals representing all the possible ages at death for a person. The size of each interval corresponds to the probability of dying at that age.
Step 3. Generate a random number uniformly distributed between 0 and 1.
Step 4. Identify the interval which contains the random number generated in Step 3. The age at death is the age which corresponds with this interval.
Method 4. Using Vitality Table with Cumulative Probabilities
Step 1. Start.
Step 2. Generate a random number uniformly distributed between 0 and 1.
Step 3. Using the Vitality table with Cumulative Probabilities table, find the age which most closely corresponds to the probability generated in Step 2. This age is the Age At Death.
Each of the above methods generates one age at death. The methods can be performed repeatedly to generate a sample set of ages at death. The sample will reflect the distribution of ages at death implied by the table used (e.g. mortality, vitality, etc.). However, any given sample may have an average or other distribution (substantially) different from that implied by the table. All of the above methods are based on random numbers uniformly distributed between 0 and 1 and may be used for randomly generating ages at death may be used for more than one person (e.g. a couple) by repeating the process for each person.
Method 5. Stratified sampling. This method provides greater assurance that the sample set of ages of death will be distributed as the mortality (vitality) table would imply. It divides the 0-1 interval into m partitions and generates S(i) random numbers [S(1) . . . S(m)] in each partition for a total sample size N=Sum[S(i)] for i=1 to m.
For each partition i=1 to m, select Si random numbers between LBOUND and UBOUND where LBOUND=Sum(Sj)/N for j=1 to i−1 (note LBOUND=0 when i=1) and UBOUND=Sum(Sj)/N for j=1 to i. Generate an Age At Death using each random number as the input in place of the random number called for in one of the 4 methods described above. This method is explained further in the following example.
Step 1. Select m as the number of partitions and Si as the number of random numbers to select in the i^{th }partition.
Step 2. Let N=Sum[S(i)] for i=1 to m
Step 3. Let PARTITION_CTR=1
Step 4. Let LBOUND=0
Step 5. Let UBOUND=S(i)
Step 6. Let CUM_SI═(PARTITION_CTR)
Step 7. Let j=1
Step 8. Let RND=a random number between 0 and 1
Step 9. Let RND=(RND*(UBOUND-LBOUND)+LBOUND)/N
Step 10. Generate an Age At Death using one of the 4 methods described above using RND in place of the random number.
Step 11. Let j=j+1
Step 12. If j≦S(PARTITION_CTR) then go to Step 8, otherwise continue
Step 13. Add 1 to PARTITION_CTR
Step 14. if PARTITION_CTR>m then Stop otherwise continue
Step 15. Let CUM_SI═CUM_SI+S(PARTITION_CTR)
Step 16. Let LBOUND=UBOUND
Step 17. Let UBOUND=CUM_SI
Step 18. Go to Step 7
The following routine generates a set of ages distributed very closely to that implied by the table. It is similar to Method 5 in which M=N and S=1. The following method generates N numbers uniformly distributed between 0 and 1. These numbers are used to generate the ages at death. It should be noted that there is no randomness implied by the following method. Rather it produces N ages at death distributed as the mortality (vitality) table would imply.
Method 6.
Step 1. Start. Select a sample size of N (e.g. 1000).
Step 2. Let a variable RND=0
Step 3. Add 1/(n+1) to RND
Step 4. Generate an Age at Death using one of the 4 methods described using RND in place of the random number.
Step 5. If RND≧n/(n+1) then stop. Otherwise return to Step 3.
The following routine generates pairs of ages at death Method 7. (Similar to Method 3a)
Method 3a. Using Mortality Table (More General Case of Method 3)
Step 1. Start
Step 2. Divide the segment 0-1 into intervals representing all the possible pairs of ages at death for a couple. The size of each interval corresponds to the probability of dying at those ages.
Step 3. Generate a random number uniformly distributed between 0 and 1.
Step 4. Identify the interval which contains the random number generated in Step 3. The ages at death are the ages which correspond to this interval.
Projections are often used in investment planning, particularly for retirement. In such a projection, the portfolio value is projected given an initial value, a series of expected cash flows, a time horizon, and a series of rates of return. Conventional methods assume a fixed time horizon and a fixed rate of return. In a retirement planning situation, the investor(s) wants to be able to project the portfolio value until his (their) death. Obviously the age at death is uncertain. It is, therefore, beneficial to calculate a number (N) of portfolio projections or trials for a set of ages at death. One technique for accomplishing this is described below.
Each trial is assigned a corresponding age at death which determines the length of time the trial will run. The age of death can be derived or calculated or selected from any of the methods disclosed above. An age at death may be derived for each member of a couple (or even a group). The trial will end when the last person in the group dies. The portfolio value is initialized to a beginning or initial value. For each period in the projection the portfolio value is updated based on a rate of return (either fixed or randomly generated). The portfolio value is adjusted for any cash flows expected to occur within that period. A trial may be terminated early (before death) if the portfolio runs out of money, or the value may be allowed to assume negative values. A trial is considered a success if the ending portfolio value exceeds some target value. Otherwise it is a failure.
Running large numbers of trials lets a user of the method derive or calculate success of a particular portfolio based on the selected variables. The distribution of portfolio values at various time periods (e.g. after 1 year, 5 years, 10 years etc.) can also be examined and a set of ages at death can be generated by running any of the above methods N times to create N projections. Alternatively, we could uses any of methods 1-4 to calculate an age at death. Then we could a projection using this age. We could repeat these steps N times. Still another alternative is to create a projection that determines the age of death as the projection is created. This requires using method 1 or 2. A loop (as described in each method) is performed that also updates the portfolio value and tests whether the person is still alive. The loop continues until the person or the last person in the couple or group dies.
It should be appreciated that although the time periods are typically yearly increments, the method could be used for other periods (e.g. months or days). Allowing the ages at death to vary lends itself well to modeling cash flows that depend upon that age at death such as life insurance and social security payments. Alternatively, cumulative tables could show a lifespan (number of years to live) instead of an age. Moreover, fractional ages can be used by interpolating values in the mortality or vitality tables.
It is contemplated that the inventive method also allows a user to specify a variety of spending rules for an investor. Illustrative examples of such spending rules include: spending 5% of the portfolio value of a previous subdivided period T_{s}; spending 3% of the average value of the portfolio value over a period of time (e.g. 5 years); or initially spending 5% of the portfolio value and then adjusting the portfolio value by the anticipated inflation rate, but not allowing the investor's spending to exceed a specified maximum percentage and/or fall below a specified minimum percentage.
It is also contemplated that the method take into account an investor's expected or anticipated inheritance by, for example, applying the mortality table to his/her benefactor in the manner described above. It is further contemplated that the incorporate the effect of taxes on the value of each portfolio. Separate portfolios for taxable, tax-deferred (e.g. traditional IRA) and tax-exempt (e.g. Roth IRA) accounts may be assumed. Furthermore, different return and risk assumptions may be specified for each portfolio. Importantly, the correlation coefficient of the returns among the asset classes (e.g. bonds, large cap stocks, small cap stocks) may also be specified by the user. For each subdivided period T_{s}, rates of return may be randomly or pseudo-randomly drawn from different distributions having an assumed mean and standard deviation. In addition, the assumed rates of return may be drawn from distributions that are correlated with each other.
For taxable accounts, the method will recognize and take into account that the realized gains, based on a user-defined cost basis, turnover and tax rates, will be taxed at capital gains rates, while income (based on yield) will be taxed at ordinary income tax rates. Similarity, withdrawals from tax deferred accounts will be taxed at ordinary income tax rates while tax-exempt accounts will not be taxed.
It is also contemplated that the method allows the user to specify dynamic return-and-risk assumptions. Dynamic return-and-risk assumptions refer to the variation of the mean and standard deviation for the rate-of-return distribution for each trial. This assumption may be based for example on the passage of time (e.g. for modeling an investor who has become more conservative over time), an event (e.g. retirement of an investor), or the portfolio value itself relative to a stated goal (e.g. the calculated return required to meet the stated goal).
It is further intended that the stored mortality table be updated at appropriate times.
It is also contemplated that the inventive method allows the projected return to be based directly on historical returns. Thus, by a database of historical returns for various asset classes the inventive method and apparatus can generate a return for a subdivided period T_{s }by randomly drawing a rate of return from a historical period and applying that rate to each investment portfolio for that subdivided period T_{s}.
It should additionally be understood that although the inventive method has been described in connection with the disclosed hardware system 10, it may also be applied to a general-purpose digital computer incorporating the appropriate hardware architecture and elements.
Thus, while there have shown and described and pointed out fundamental novel features of the invention as applied to preferred embodiments thereof, it will be understood that various omissions and substitutions and changes in the form and details of the methods described and devices illustrated, and in their operation, may be made by those skilled in the art without departing from the spirit of the invention. For example, it is expressly intended that all combinations of those elements and/or method steps which perform substantially the same function in substantially the same way to achieve the same results are within the scope of the invention.
Moreover, it should be recognized that structures and/or elements and/or method steps shown and/or described in connection with any disclosed form or embodiment of the invention may be incorporated in any other disclosed or described or suggested form or embodiment as a general matter or design choice.
Cited Patent | Filing date | Publication date | Applicant | Title |
---|---|---|---|---|
US5148365 * | Aug 15, 1989 | Sep 15, 1992 | Dembo Ron S | Scenario optimization |
US5754980 | May 24, 1995 | May 19, 1998 | Century Associates L.L.C. | Method of providing for a future benefit conditioned on life expectancies of both an insured and a beneficiary |
US5799287 * | May 30, 1997 | Aug 25, 1998 | Dembo; Ron S. | Method and apparatus for optimal portfolio replication |
US5806047 * | Sep 25, 1995 | Sep 8, 1998 | Metlife Life Insurance Company | System for combined pool portfolio |
US5812987 * | Mar 13, 1995 | Sep 22, 1998 | Barclays Global Investors, National Association | Investment fund management method and system with dynamic risk adjusted allocation of assets |
US5819238 * | Dec 13, 1996 | Oct 6, 1998 | Enhanced Investment Technologies, Inc. | Apparatus and accompanying methods for automatically modifying a financial portfolio through dynamic re-weighting based on a non-constant function of current capitalization weights |
US5857176 * | May 9, 1997 | Jan 5, 1999 | Cantor Fitzgerald & Co., Inc. | Fixed income portfolio data processor |
US5864828 * | Aug 27, 1991 | Jan 26, 1999 | Proprietary Financial Products, Inc. | Personal financial management system for creation of a client portfolio of investment and credit facilities where funds are distributed based on a preferred allocation |
US5878405 * | Sep 25, 1996 | Mar 2, 1999 | Coordinated Data Services, Inc. | Pension planning and liquidity management system |
US5884287 * | Apr 11, 1997 | Mar 16, 1999 | Lfg, Inc. | System and method for generating and displaying risk and return in an investment portfolio |
US5913198 * | Sep 9, 1997 | Jun 15, 1999 | Sbp Services, Inc. | System and method for designing and administering survivor benefit plans |
US5918217 * | Dec 10, 1997 | Jun 29, 1999 | Financial Engines, Inc. | User interface for a financial advisory system |
US5930760 * | May 1, 1997 | Jul 27, 1999 | Impaired Life Services Limited | Process and apparatus for the derivation of annuity rates |
US5956691 * | Jan 7, 1997 | Sep 21, 1999 | Second Opinion Financial Systems, Inc. | Dynamic policy illustration system |
US6012043 * | Sep 9, 1996 | Jan 4, 2000 | Nationwide Mutual Insurance Co. | Computerized system and method used in financial planning |
US6012044 * | May 25, 1999 | Jan 4, 2000 | Financial Engines, Inc. | User interface for a financial advisory system |
US6064969 * | Sep 8, 1995 | May 16, 2000 | Citicorp Life Insurance Company | Flexible annuity settlement proposal generating system |
US6615180 * | Mar 19, 1999 | Sep 2, 2003 | Impaired Life Services Limited | Process and apparatus for the derivation of annuity rates |
US20040177021 * | Mar 3, 2004 | Sep 9, 2004 | Carlson Joseph W. | Apparatus and method for achieving enhanced returns on investments |
US20050027645 * | Jan 31, 2003 | Feb 3, 2005 | Wai Shing Lui William | Business enterprise risk model and method |
WO1998044444A1 * | Apr 2, 1998 | Oct 8, 1998 | William J Fox | Method and apparatus for virtual investment advisor and support system |
WO1999028845A1 | Sep 23, 1998 | Jun 10, 1999 | Geert Bekaert | Financial advisory system |
Reference | ||
---|---|---|
1 | * | Leimer, Dean "A guide to social security money's worth issues" Social Security Bulletin, vol. 58 n2, Summer 1995. |
Citing Patent | Filing date | Publication date | Applicant | Title |
---|---|---|---|---|
US7962388 * | Jun 3, 2008 | Jun 14, 2011 | Macey-Holland & Co., Llc | System and method for incorporating mortality risk in an investment planning model |
US7962392 * | Sep 24, 2009 | Jun 14, 2011 | Macey-Holland & Co., Llc | System and method for incorporating mortality risk in an investment planning model |
US8175900 * | Mar 2, 2009 | May 8, 2012 | The Trustee and Successor Trustees of RGD 2006 Trust | Life insurance strategic value |
US8566196 * | Jun 13, 2011 | Oct 22, 2013 | Macey-Holland & Company, LLC | System and method for incorporating mortality risk in an investment planning model |
US8635132 * | Apr 14, 2008 | Jan 21, 2014 | United Services Automobile Associatiion (USAA) | Self-service real-time financial advice |
US20080243665 * | Jun 3, 2008 | Oct 2, 2008 | Rex Macey | System and method for incorporating mortality risk in an investment planning model |
US20100076905 * | Sep 24, 2009 | Mar 25, 2010 | Rex Macey | System and method for incorporating mortality risk in an investment planning model |
US20100223077 * | Mar 2, 2009 | Sep 2, 2010 | The Trustee and Successor Trustees of RGD 2006 Trust | Life insurance strategic value |
US20110276517 * | Nov 10, 2011 | Rex Macey | System and method for incorporating mortality risk in an investment planning model |
U.S. Classification | 705/35, 705/4, 705/36.00R, 705/7.11 |
International Classification | G06Q40/00, G06Q10/06, G06F17/30 |
Cooperative Classification | G06Q40/06, G06Q10/063, G06Q40/00, G06Q40/08 |
European Classification | G06Q40/06, G06Q40/08, G06Q40/00, G06Q10/063 |
Date | Code | Event | Description |
---|---|---|---|
May 24, 2013 | REMI | Maintenance fee reminder mailed | |
Aug 16, 2013 | FPAY | Fee payment | Year of fee payment: 4 |
Aug 16, 2013 | SULP | Surcharge for late payment |